v0.8.2

[qinxuye]

This is the release notes of v0.8.2. See here for the complete list of solved issues and merged PRs.

New Features

• DataFrame
  • Support inclusive argument for pd.date_range ([BACKPORT] Support inclusive argument for pd.date_range (#2718) #2721)

Enhancements

• Optimize eval-setitem expressions as single eval expressions ([BACKPORT] Optimize eval-setitem expressions as single eval expressions (#2695) #2699)
• [Ray] Refine raydataset integration ([BACKPORT][Ray] Refine raydataset integration (#2579) #2712)
• [Ray] refine ray dataset integration ([BACKPORT][Ray] refine ray dataset integration (#2705) #2726)
• Add support for reading partitioned parquet for fastparquet ([BACKPORT] Add support for reading partitioned parquet for fastparquet (#2724) #2729)
• Fix duplicate exceptions in log ([BACKPORT] Fix duplicate exceptions in log (#2723) #2736)

Bug fixes

• Fix sort_values for empty DataFrame or Series ([BACKPORT] Fix sort_values for empty DataFrame or Series (#2681) #2686)
• Eliminate redundant eval node in optimization ([BACKPORT] Eliminate redundant eval node in optimization (#2683) #2688)
• Avoid iterative tiling for df.loc[:, fields] ([BACKPORT] Avoid iterative tiling for df.loc[:, fields] (#2685) #2689)
• Fix use_arrow_dtype parameter for read_parquet ([BACKPORT] Fix use_arrow_dtype parameter for read_parquet (#2698) #2702)
• Fix error on dependent DataFrame setitems ([BACKPORT] Fix error on dependent DataFrame setitems (#2701) #2703)
• Fix estimate_pandas_size on pd.MultiIndex ([BACKPORT] Fix estimate_pandas_size on pd.MultiIndex (#2707) #2710)
• Import vineyard.data.pickle to make members available ([BACKPORT] Import vineyard.data.pickle to make members available (#2714) #2716)
• Fix shuffle when ndim of input tensors are different ([BACKPORT] Fix shuffle when ndim of input tensors are different (#2727) #2728)

---

This discussion was created from the release v0.8.2.